Last April I devised a trading experiment that seemed destined to work out. MB Trading offers a unique pay for limit order program that I first covered a year and a half ago. The goal was to suck commissions from the market by trading with limit orders as frequently as possible.
Zero expectation
The ideal strategy is one that consistently anticipates the correct market direction and that does not lose more than it makes. I don’t happen to have an off the shelf strategy that would pair nicely with limit orders. Instead, I opted for a strategy that’s supposed to come out at exactly break even.
The idea is that if we could trade with limit orders as frequently as possible, then a normal strategy is not required. The commissions earned would act like trading profits, while the strategy ignores the results from changes in the market price. The approach should also earn the spread as a result of making a market.
I accomplished this by arbitrarily choosing fixed pip distances to test on the EURUSD. A 20 pip stop loss struck me as frequent enough to facilitate 100+ trades per day without affecting the execution. All trades entered via limit order to earn the $1.95 per 100k traded. The strategy goal was to maximize the number of trades that exited via limit order (i.e., the take profit) and to limit the number of exits via stop loss.
The closer that the take profit appears to the entry price, the more probable that the trade would hit its take profit. I made a table below to illustrate the concept.
| Take Profit | Stop Loss | Percent Wins | Value of Wins | Percent Losses | Value of Losses | Net Outcome |
|---|---|---|---|---|---|---|
| 3 | 20 | 20 / (20 + 3 ) = 86.956% | 86.956 * 3 = 260.87 | 3 / (20 + 3) = 13.043% | 13.043 * 20 = 260.87 | 260.87 – 260.87 = 0 |
| 7 | 20 | 20 / (20 + 7) = 74.074% | 74.074 * 7 = 518.52 | 100%-74.074%= 25.926% | 25.926 * 20 = 518.52 | 518.52-518.52 = 0 |
| 10 | 20 | 20 / (20+10) = 66.667% | 66.667 * 10 = 666.7 | 100%-66.667% = 33.333% | 33.333 * 20 = 666.7 | 666.7 – 666.7 = 0 |
| 15 | 20 | 20 / (15+20) = 57.143% | 57.143 * 15 = 857.143 | 100% – 57.143% = 42.857% | 42.857 * 20 = 857.143 | 857.143-857.143 = 0 |
Notice how the results have nothing to do with markets. It’s basic probability, assuming that markets follow an independent distribution of outcomes.
Expected Earnings from Commissions
Hopefully, you are convinced that trading with fixed stop losses and take profits and no indication of direction is likely to lead to exactly no profit and loss. My goal now is to show you how trading at random this way could potentially lead to profits.
The MB Trading scheme is that they pay $19.95 per million for limit orders. Stop and market orders pay $29.95 per million. That works out to $0.01995 per microlot for limits and -$0.02995 per microlot for other orders. Let’s break this down into our expected profit and loss after 100 micro lot trades.
All orders enter the market using a limit order. That means that all 100 trades will earn the commission upon entry. The value of trade entries is 100 trades * $0.01995/trade = $1.995.
The remaining profit and loss comes from exiting trades. Assuming that we use a take profit of 3 and a stop loss of 20, then we expect to win 86.956% of the time and to lose 13.043%.
We expect 86.956 trades to earn $0.01995. 86.956 trades * $0.01995 = $1.735.
We expect 13.043 trades to lose $0.02995. 13.043 * -$0.02995 = -$0.391
The value of the trade exits is $1.735 – $0.391 = $1.344
The value of all trades is the sum of the profits from the entries and exits. $1.995 + $1.334 = $3.339 after 100 microlot trades. All that needs to happen now is for the breakeven strategy described above to not lose, or at least to not lose more than 33 pips.
You’ll notice that these numbers differ slightly from the video. That’s because I elected to assume that the base currency in our forex trades was USD. If we used EUR when EURUSD traded at 1.30, then the numbers would match what you see in the video.
Changed probabilities
I discovered after hundreds of completed trades that the anticipated probabilities were always several percentage points worse than expected. I noticed from watching the strategy that maintaining a fixed price for the entry allowed the spread to compress. That’s great for MB Trading and the people on the other side of my trade. What did not work for me was that the forex market depth would drift away from the entry price.
When someone finally accepted by bid or offer, the rest of the liquidity would be several micropips or even a pip beyond where I got executed. Entering via limit at best bid/best offer appears to “give up” some of the potential profit. The logical next step is to place the orders deeper in the book to make sure that this “give up” is minimized.
One day , i think you will win a Nobel Prize for something ,heavens knows what!! Its fascinating watching the process, like a form of mental gymnastics! good work cheers Baz
Haha, thanks for the encouragement. I’m glad you enjoyed the post. When I began the test, I always figured that I’d at least get some blogging mileage out of it if the strategy flopped.